Hello,

I'm trying to simulate 20MHz crystal oscillator in Ltspice.
I found crystal oscillator has the same circuit with RLC, however I don't know how to select RLC value using the datasheet.
Can you let me know how to select the values with the reliable formula?
>Datasheet
View attachment 325491
>Circuit for oscillator
View attachment 325492

 

Hi jae.
Welcome to AAC.
Your links report Error 404, Not Found.
E

 

hi jae,
You may find this method helpful.
The parameters are for a 4MHz xtal.
Fine tune the parameters from the 20Mhz datasheet.

E

 

hi jae,
You may find this method helpful.
The parameters are for a 4MHz xtal.
Fine tune the parameters from the 20Mhz datasheet.

E
View attachment 325494

Hello ericgibbs,
I am so sorry for the link is unavailable to access, so I attach my osc spec again!

Your attached file help me a lot, but I have still problem to set up the values in the attached simulation!

 

This is only an estimate of 20MHz crystal
Motional parameters for an equivalent circuit
Rs=12Ω Cp=6.3pF Ls= 2.49mH Cs=25.3fF

 

Hi jae,
This is the best approximation I can do, based on the limited information in your datasheet.

E

 

I like the way Eric uses simulation, very valuable simulation. Thanks, you and many others with different talents.
Using a Gilbert cell with a crystal The NE602 was popular during that era at signetics when high tech had to rely on crystals.
An article shows variety series and parallel oscillators. I posted parameters that are not so good but are helpful to understanding
beyond parameters, understanding which topology or configuration does what is does is important. Don't stress let the NE602 perform.

I have a number of different test boards. I plug the crystal into different boards, one board has sockets for trying different capacitors.
One board can adjust the gain. You can always lower the drive power. One is series one is parallel useful in fine tuning parameters.
The circuit in fig 9 the author might want to know how much capacitance it will take to pull frequency for a given drive power.
Knowing when the oscillation starts to dampen. It has a differential amplifier and Gilbert Cell also has ability to mix precisely
outputting standard impedance is where the product is filtered for amplification and use. Doing that with 2 transistor stages?

There are many other effects that will help you to design a better than average dedicated crystal oscillator circuit.
I think it started out crystal oscillator types began to refine. Butler type were found to propagate symmetric sinewave,
Colpitts maybe reliability, Pierce frequency accuracy. In this article the combination of Local Oscillator and Gilbert Cell simplifies
frequency conversion which is either a blessing or a curse seen on Spec analyzer It depends on your skills and learning curve.
If you mix it is balanced if you don't that's ok. The other useful frequencies are usually in addition to a fundamental.
Circuits - Single-Chip Frequency Converter (n5dux.com)

 

Here is an additional example to show how the motional parameters are computed from frequency, Q, and series resistance in the explicit subcircuit for the crystal.

 

Development of simulation, Whats your favorite method includes simulation and actual how to measure a crystals motional parameters.
Demonstration notes by W2AEW of a circuit that could be simulated for educational reasons to understand better.
The simulation is ideal, the NanoVNA methods for crystals are getting better while passing those parameters back and forth.
You have 3 components: the test board, the crystal and the test instrument. We reducing error using proven techniques.
The test board capacitance accounted for in simulation and measured using a capacitance meter shown here in real test.

The development history of spice crystal parameters, Not sure?, I remember I needed to buy a filter software, because an excel
spreadsheet was included that calculated crystal motional parameters. It was confusing but it worked. Don't remember the cost.
Eric and others have donated LTspice simulations, a whole lot easier and there is no cost. Again, thanks to the many people that
volunteered so much effort into producing these excellent bode plots showing parallel and series resonance and Q.
Also, those video demonstrations in simple terms how it is done. The efforts using NanoVNA. LtSpice.

W2AEW on "How to use G3UUR method and NanoVNA" cited reference to K8ZOA comparison,
W2AEW instrumentation setup, bnc sockets cables, test socket and prototype board on copper clad. Need to be very careful.
The capability of determining the actual crystal under test and what it involves, comparison of methods.
Page 3 Measuring the "holder" with capacitance meters
Page 4 The series-thru VNA test fixture
Page 5 Switches from ground capacitor to ground only. This gives a difference in frequency measured on the frequency counter.
Then the 3 methods using NanoVNA.
crystal_measurements_nanovna.pdf (qsl.net)